This proposal is a continuation of a study examining expressed emotion (EE) in the families of schizophrenic patients who are participating in a two-year controlled trial comparing two pharmacologic treatment strategies - maintenance neuroleptics versus a "targeted" strategy in which drugs are used only to treat an emergent psychotic episode. The Camberwell Family Interview is used to assess EE. The hypotheses being tested are 1). high EE will correlate with poor outcome in patients over the two-year period; 2.) patients from low EE families are expected to do equally well whether randomized to targeted or maintenance drugs, whereas patients from high EE families will have a much poorer outcome if not assigned to maintenance drugs. Content analysis of brief speech samples will be assessed as simplified alternatives for rating EE. Furthermore, patients' attitudes toward family members, as well as their perception of their relatives' attitudes of criticism and hostility directed toward themselves, will be evaluated as alternative variables (possibly superior to EE) for predicting patient outcome and need for maintenance antipsychotics. Changes in EE over time in response to treatment will be examined. Attributes of this sample such as race, socioeconomic status, and pattern of illness will be considered as possible modifiers of the impact of EE on patient outcome. The identification of family attributes that relate to patient outcome would provide a focus for therapeutic intervention with families. Relating such factors to the need for medication could identify a group of patients for whom the risks of continuous neuroleptic exposure might be unnecessary. Simplified instruments would allow the application of such findings to a wide range of clinical settings.